1. Field of the Invention
The present invention relates to scanning probe microscopes (SPMs) and, more particularly, to improved SPMs capable of referencing height measurements to the sample surface in a variety of applications.
2. Description of the Related Art
An SPM is capable of high-resolution non-destructive measurements in a variety of semiconductor applications. The atomic force microscope (AFM) is the SPM most extensively used today for imaging and characterizing step height and surface roughness on samples such as silicon wafers and thin film magnetic heads. It has been referred to as an essential characterization tool (see, for example, Strausser, Schroth, and Sweeny, "Characterization of the LPCVD Grown Rugged Polysilicon Surface Using Atomic Force Microscopy," Future Fab International, Issue 4, Vol. 1, 307-311 (1997)). The typical AFM includes a probe which includes a flexible cantilever and a tip mounted on the free end of the cantilever. The probe is mounted on a measurement head that is mounted on a common mechanical reference structure with the sample. The mechanical head typically includes an XY actuator assembly and a Z actuator. The XY actuator assembly drives the probe to move in an X-Y plane for scanning. The Z actuator, which is mounted on the XY actuator assembly and which supports the probe, drives the probe to move in a Z axis extending orthogonally to the X-Y plane.
AFMs can be operated in different modes including contact mode and TappingMode. In contact mode, the cantilever is not oscillated, and cantilever deflection is monitored as the tip is dragged over the sample surface. In TappingMode (Tapping and TappingMode are trademarks of Veeco Instruments, Inc.), the cantilever is oscillated mechanically at or near its resonant frequency so that the tip repeatedly taps the sample surface or otherwise interacts with it, dissipating energy and reducing the cantilever's oscillation amplitude. The oscillation amplitude indicates proximity to the surface and may be used as a signal for feedback to reduce the forces between the probe-tip and the sample surface, and thus prevent damage or contamination that may be caused by a contact-mode measurement. In fact, in many cases it may be necessary for the reference measurement to be a non-contact measurement. In any operational mode, interaction between the probe and the sample surface induces a discernable effect on a probe operational parameter, such as cantilever deflection, that is detectable by a sensor.
Next, note that wafer scale SPM's have lateral and vertical resolutions of roughly 2 nm and 0.1 nm, respectively. The 0.1 nm vertical resolution is typically limited to vibrational and thermal noise between the associated probe tip and the sample. Nevertheless, the SPM can obtain detailed topographic information in ambient air. Surface images of this kind are not available with the scanning electron microscope (SEM) because of its relatively large depth of focus. Surface details at this level can be resolved in the transmission electron microscope (TEM), but the procedure requires sectioned samples, and is tedious, time consuming, and destructive.
For both step height and surface roughness measurements, the SPM's utility is in its metrological capabilities and not its imaging capability--meaning that the numerical values produced are more important than visualization of the surface. Importantly, features on samples measured by a SPM often need to be resolved beyond the 0.1 nm limitation. Many different types of SPM sensors, with sensitivities beyond 0.1 nm, have been proposed. These include: 1) the common optical-lever sensor (see, e.g., Meyer and Amer, "Novel Optical Approach to Atomic Force Microscopy," Appl. Phys. Lett. 53, 1045 (1988); Alexander, Hellemans, Marti, Schneir, Elings, Hansma, Longmire, and Gurley, "An Atomic-Resolution Atomic-Force Microscope Implemented Using an Optical Lever," Appl. Phys. Lett. 65 164 (1989); 2) the piezoresistive sensor (Tortonese, Barrett, and Quate, "Atomic Resolution With an Atomic Force Microscope Using Piezoresistive Detection," Appl. Phys. Lett., 62, 8, 834-6 (1993)); 3) and the interdigital sensor (Manalis, Minne, Atalar, and Quate, "Interdigital Cantilevers for Atomic Force Microscopy," Appl. Phys. Lett., 69 (25) 3944-6 (1996); Yoralioglu, Atalar, Manalis, and Quate, "Analysis and design of an interdigital cantilever as a displacement sensor," 83(12) 7405 (June 1998)). Unfortunately, these sensors have shown ultra-low sensitivities only when either not in contact with the sample, or when using a highly specialized and expensive apparatus, which is not suitable for a manufacturing environment.
Practical limitations restrict the typical SPM when operating over large areas at high speeds. The area restriction is primarily due to the lack of low-noise, large displacement scanners. Scanners that can traverse larger areas have greater vibration noise. In light of the fact that the SPM's use is already limited due to drawbacks associated with vibrations in the operating environment, any increase in noise due to scanner operation significantly exacerbates the SPM's performance problems.
Many applications in the microelectronics industry need the resolution of the SPM over large areas. As features on microelectronics and data storage devices continue to shrink, this need will only increase further. In fact, features of interest are quickly falling below the optical diffraction limit and the optical scatter detection limit of the instrument. This means tools that rely on light may no longer be able to service the industry in high throughput large area defect analysis.
In addition, AFMs have historically had problems with thermal drift and acoustic vibrations in the mechanical path between the probe and the sample. Changes in the mechanical reference path appear as noise in the AFM topography image. In current AFM designs, this noise is reduced by making the mechanical reference path rigid and of thermally stable materials that are relatively expensive and that do not eliminate thermal drift and acoustic vibrations. A more cost effective and reliable solution is desired.
One solution in such a case is to make a differential AFM measurement optically. However, such a solution is inadequate for applications contemplated by the present invention. In particular, optical displacement measurements are problematic on unknown sample surfaces because, for example, topographic features can cause scattering of the reflected light which can yield erroneous data.
The art is in need of an improved SPM that can operate over large areas at high speeds for both metrology and defect analysis while minimizing the effects of noise in the mechanical reference path. The solution should be applicable to the standard SPM to increase sensitivity for defect review and surface visualization. Furthermore, the scanner vibration restrictions for both systems should be significantly reduced and, in certain applications, eliminated.